Page 2 of 2

Posted: Fri Jun 18, 2004 5:21 pm
by blueznl
mmm why would it be confusing? it's rather simple: use the last type def the compiler encountered

if one adds the type def to the pointer, from that moment on the def changes

i agree that if a change would be temporary the use of parenthesis would be better

*a.(lasjdf)\asdf = asdfasdf

but it will raise quite a few questions with the crowd... i'd suggest to move the last parenthesis:

*a.(asdf\asdf) = asdfasd

or perhaps use a different type of bracket, dunno

*a.[asdf\asdf] = asdfasd

square brackets have a slight preference for me, as in (most) basics normal parenthesis are 'enclosures' of functions or expressions, but not an inherent part of the syntax, they might cause some confusion with dimmed stuff...

dim *a.b(100)

*a(x) = 5
*a.(x)\b = 5

while square brackets in a different place are slightly easier to read:

*a.[x\b] = 5

Posted: Fri Jun 18, 2004 5:46 pm
by PolyVector
I agree with Fred on the more c-like implimentation for temperary use...

If there needs to be a change-as-you-go method, then I say keep the current syntax, and allow the type to be changed with another standard definition:

Code: Select all

*Var.StructABC\a=1
*Var\b=2
*Var\c=3

*Var2.StructXYZ\x=4
*Var2.(StructABC)\b=5
*Var2\z=6

Posted: Mon Aug 30, 2004 8:55 am
by soerenkj
is this going to be implemented..?
I really need it now!